Correspondence¹ re: S. Shimoyama et al., Increased Serum Angiogenin Concentration in Colorectal Cancer Is Correlated with Cancer Progression. Clin. Cancer Res., 5: 1125–1130, 1999.

In September 1998, we published in Clinical Cancer Research a study examining the clinical value of angiogenin, a protein implicated in angiogenesis (1). Our article described the analysis of breast carcinoma specimens for the expression of angiogenin. We showed that angiogenin expression was elevated in a majority of breast carcinoma extracts and that the expression of tumor angiogenin, but not of circulating angiogenin, was an independent prognostic indicator of favorable outcome in patients with operable breast cancer (2). Recently, Shimoyama et al. (3) have reported that serum angiogenin concentration is elevated in colorectal patients and that the degree of its elevation correlates with the degree of cancer progression. However, these authors do not demonstrate the reverse correlation between circulating angiogenin levels and survival, and they do not elucidate the source of angiogenin in the colorectal samples. In our clinical study, the issue of whether the angiogenin was produced in the breast carcinoma cells or in the surrounding stroma was not addressed. To establish the cellular source of angiogenin in breast cancer tissue, we performed additional immunohistochemical analysis.

Tissue sections were obtained from paraffin blocks of breast carcinomas in which cytosol angiogenin had been determined previously. In addition, samples of uninvolved breast were acquired. Incubation with a polyclonal anti-angiogenin antibody (R&D Systems) diluted at 1:20 in PBS (pH 7.4) was done on paraffin material. Sections were stained by avidin-biotin complex immunoperoxidase method and counterstained with hematoxilyn to visualize cell nuclei. The negative controls used were sections from the same tissue block in which the primary antibody was omitted. A cellular qualitative analysis of angiogenin staining was done. Sections were interpreted blinded to the previous angiogenin expression assays and to the clinical results.

In the breast carcinoma, the staining pattern for angiogenin was homogeneous and showed a strong cytoplasmic immunoreactivity located on the tumor cells (Fig. 1, left). There were only a small number of weakly staining stromal cells such as fibroblasts, inflammatory cells, and capillary endothelial cells in the vicinity of the malignant cells. The epithelial localization of angiogenin was confirmed in normal breast specimens. Angiogenin immunodetection showed a staining pattern that was confined to the cytoplasm of the normal ductal epithelial cells (Fig. 1, right).

Our result demonstrates that angiogenin expression in breast cancer is mainly located in the carcinoma cells. Although angiogenin protein expression has not been studied before in breast tumors, our immunohistochemical data are in accordance with mRNA localization obtained by in situ hibridization showing the presence of the angiogenin in other malignant cells such as colonic, gastric, and hepatocellular carcinoma cells (4). In normal breast specimens, angiogenin staining was confined to the epithelial cells of ducts. The presence of angiogenin within neoplastic and normal breast epithelium indicates that angiogenin is synthesized by breast cells in vivo. We reported previously that angiogenin expression had a favorable impact in breast cancer patient prognosis and that angiogenin is a substrate for the adhesion of breast cancer cells in vitro (2, 5). Taken together, these data suggest that angiogenin plays a role in breast cancer progression in process that is related to cellular adhesion.